The Metis project

Dependable cooperative systems for maritime safety and security

Efficient monitoring and management of maritime activities is a critical task for all coastal states. It is necessary for collision avoidance, enforcement of fishing policies, pollution control, deterring criminal activities such as smuggling & terrorism and guidance in cases of bad weather or unexpected obstacles. However, as the volume of transported goods grows and the range of activities expands, so too do the risks and impact of accidents, pollution and criminal activities. The challenge today is how to effectively monitor and manage coastal areas with a minimum of resources in terms of skilled manpower and active enforcement.

A coastal area being monitored:

Coastal area monitoredClick to enlarge


monitoring coastal areaClick to go to the video


Environmental hazardClick to enlarge

Research results

To drive the application domain for dependable cooperative systems to a higher level, the demanding domain of maritime safety and security has been used. Here many vessels at sea are monitored for position, behaviour, recent history and identity. This directly assists the authorities to reduce the number of accidents and to expose potential criminal activities.

The main results of the project include:Metis illustration

  • Technology for extracting key information from unstructured data sources
  • Temporal records matching
  • Advanced probabilistic reasoning using “aged” input data
  • An novel intuitive visual representation of decision making rationale
  • Glyphs to provide a visual representation of complex and uncertain information

Value proposition

The technologies developed by the Metis project are general in nature and can be applied to other domains requiring intelligent data collection & assimilation, advanced reasoning, configuration optimization and intuitive presentation of findings. In addition to a variety of safety and security cases other domains can benefit. A typical example being Condition Based Maintenance, here the data coming from many sensors monitoring multiple parameters within a complex system can be reasoned upon and a prognosis given indicating the likelihood of failure. With this information, preventative maintenance can be carried out before any failure actually occurs. Such additional functionally can bring significant economic benefits to owners and users.